ALMA-IMF Core Working Group, October 15 2020

Connection:
https://ufl.zoom.us/j/96368629056?pwd=c1JmbWN6Rll0ZkNjYzJZamtXZFdmUT09

Attendees: Fred, Yohan, Adam, Timea, Sasha, Gemma, Allison, Hongli, Roberto, Alex, Ant, Nathalie, Amy, Fabien, Tapas
Not attending: Manuel

For memory:
Minutes from our last telecon
Members and their main interests


Agenda:

1. Temperature estimates with PPMAP

• The PPMAP procedure to estimate the Tdust image of ALMA-IMF fields
  

Alex’s slides

• Herschel-only (14’’-resolution, no filtering) run
  

Bands (in micron): 70, 160, 250, 350 (Herschel), 500
→ to be used as a model toward which the high-resolution Tdust images should converge

• Run with 60’’-filtered maps (2.5’’-resolution, B3 filtering, without Herschel 500)
  

Bands (in micron): 70, 160, 250, 350 (SABOCA), 870, 1300, 3000

• Many pixels are not fitted because of negative levels created by the filtering process or because of noise in the ALMA (especially B3) images. → Lots of/Too many NANs.
  
• The SED fits of the main peaks (MM2 and MM3) still diverge at the LABOCA 870 and ALMA 1300 and 3000 micron wavelengths. → Change the weights?
  
• The ‘half noise’ and ‘tight noise’ runs try to compensate for the poor noise levels of the submm images: 
  

Half noise: sigma/2 for LABOCA (870), ALMA B6 (1300), and ALMA B3 (3000).
Tight noise: sigma/2 for LABOCA (870) and ALMA B3 (3000); sigma/4 for ALMA B6 (1300).
Gaussian noise is assumed and levels are calculated for each map as a first step of PPMAP. These noise levels are used in the SED fit. 
The effect of a non-Gaussian noise on the PPMAP results should be checked… If we were to impose some weights for the different wavelengths, PPMAP would need to be modified. → not for now!
Temperature increases from the ‘normal noise’ to ‘half noise’ and the ‘tight noise’ runs. Be careful! The “tight noise” run gives temperatures closer to that of the Herschel run. 

• Run with 30’’-filtered maps (2.5’’-resolution, B6 filtering, without SABOCA?)
  

Bands (in micron): 70, 160, 250, 350 (SABOCA), 870, 1300, 3000
The SED fit gets worse to the 70 micron point and a s a consequence the small heating toward MM2 is lost → the 60’’-filtering run seems more reasonable

• Run without the B3 data
  

Bands (in micron): 70, 160, 250, 350 (SABOCA), 870, 1300

• When the ‘free-free-contaminated’ B3 image is ignored, the Tdust image displays less cold values at the location of the HII region (18K instead of 13K) and of the filament (23K instead of 18K) lying below the HII region. → To be checked if this map would be better to estimate the temperature of cores lying over this filament…
  
• There are less NANs suggesting that a SED fit should be allowed when either the B6 or B3 image has negative pixels.
  

Preliminary conclusions:
→ The 60’’-filtering run seems better than the 30’’-filtering run
→ The ‘tight noise’ run gives temperature in the background that better corresponds to that of the Herschel run.
→ The run without B3 gives better Tdust values close to/but off the HII region.
→ Could we combine the Herschel-only map and 60’’-filtering ‘tight-noise’ map to have a continuous Tdust image? Could we fit SEDs on each pixel just dropping the negative flux?

• Jeff’s evaluation of the filtering scales of the datasets used by PPMAP 
  

Jeff’s presentation from 2 weeks ago.
Not discussed today

1. News on core extraction by Yohan and Fabien

Not discussed today

1. Other estimations of the core temperature

Not discussed today

• Status of Brian’s efforts to analyze and collect ammonia data
  

• Fred and Nathalie should soon present the proxy proposed for the protostellar luminosity of continuum cores, using the “line contamination” of “the continuum band” by COMs lines.